Immunology and Cell Biology (2013) 91, 661–664 & 2013 Australasian Society for Immunology Inc. All rights reserved 0818-9641/13 www.nature.com/icb
SHORT COMMUNICATION
Progranulin does not inhibit TNF and lymphotoxin-a signalling through TNF receptor 1 Nima Etemadi1,2, Andrew Webb1,2, Aleksandra Bankovacki1,2, John Silke1,2 and Ueli Nachbur1,2 Progranulin (proepithelin, granulin precursor) has been recently suggested to exhibit anti-inflammatory properties by directly binding to tumour necrosis factor (TNF) receptors and thereby inhibiting TNF signalling by Tang et al. This finding was challenged by Chen et al. and no interaction between progranulin and TNF receptor (TNFR) 1 or 2 was observed. We tested the ability of recombinant progranulin from different commercial sources to inhibit TNF- or lymphotoxin-a-induced signalling through TNFR1. We observed that progranulin does not affect signalling and cell death induction downstream of TNF or lymphotoxin-a. Our results suggest that the anti-inflammatory role of progranulin is not mediated through direct inhibition of TNFR1. Immunology and Cell Biology (2013) 91, 661–664; doi:10.1038/icb.2013.53; published online 8 October 2013 Keywords: cell death; inflammation; MAPK; NFkB; progranulin; TNF receptor 1
Progranulin (PGRN) is a growth factor initially isolated from supernatants of the teratoma PC cell line.1,2 It is secreted as a fulllength protein and undergoes proteolysis to release small granulins. PGRN is implicated in different physiological and pathological processes including wound repair, host defence, inflammation, tumorigenesis and neurodegeneration.3,4 Several studies have shown that PGRN has anti-inflammatory effects,5–7 and Grn / mice display increased neuroinflammation upon injury in the central nervous system8 and increased susceptibility to arthritis,9 dermatitis10 and atherosclerosis11 due to enhanced inflammatory responses. Tang et al.9 recently reported that recombinant PGRN or Atsttrin (a synthetic protein composed of three granulin fragments) rescued the inflammatory phenotype in a collagen-induced arthritis model and reduced the severity of dermatitis in an oxazolone-induced dermatitis model.10 These observations were attributed to PGRN or Atsttrin directly binding to tumour necrosis factor (TNF) receptors and blocking TNF signalling, which would be consistent with the known role of TNF as a master inflammatory cytokine.12 The ability of PGRN to act as a TNF receptor antagonist was, however, recently challenged by Chen et al.,13 who showed that PGRN does not bind to TNF receptors and therefore does not directly affect TNF signalling. To provide some perspective to this controversy, we would like to report our own results where, stimulated by the Tang et al. publication, we sought to test the hypothesis that PGRN could inhibit TNF/TNFR1 signalling differently to lymphotoxin-a (LTa)/TNFR1 signalling.
RESULTS Neither TNF nor LTa induced activation of NF-jB and mitogen-activated protein kinases is inhibited by PGRN To test our hypothesis, we examined the ability of PGRN to inhibit activation of NF-kB and mitogen-activated protein kinases by either TNF or LTa, two ligands that bind and signal via TNFR1.14 To test the reported functionality of PGRN, we treated wild-type bone marrowderived macrophages (BMDMs) for up to 2 h with TNF or LTa in the presence or absence of recombinant human PGRN (Figure 1a). Analysis of well-described signalling outcomes upon TNF or LTa stimulation, including extracellular-signal-related-kinse (ERK), c-Jun N-terminal kinase (JNK) and p38 phosphorylation and IkBa degradation, did not reveal any inhibition by PGRN. Because this result did not agree with the result described by Tang et al. we tested human and mouse PGRN from a different commercial source (Adipogen), which was reported to be a reliable source for functional PGRN.15 However, regardless of the source or the species of PGRN, we did not observe any inhibition of TNF-induced ERK phosphorylation or IkBa degradation (Figure 1b). In contrast, a TNFR1blocking antibody was very efficient at blocking these pathways showing that this assay accurately reported on TNFR1-dependent signalling pathways. Although these initial results excluded a role of PGRN on transient TNF signalling, it was possible that PGRN affected TNFR1 signalling only over a longer time period. To test this hypothesis, we examined the long-term effects of all three PGRNs on the human leukemic cell line U937, which was stably transfected
1CSCD Division, The Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia and 2Department of Medical Biology, University of Melbourne, Parkville, Victoria, Australia Correspondence: Dr U Nachbur, CSCD Division, Level 5 West, The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Victoria 3052, Australia. E-mail:
[email protected] Received 23 July 2013; revised 29 August 2013; accepted 3 September 2013; published online 8 October 2013
Progranulin does not inhibit TNFR1 signalling N Etemadi et al 662
Figure 1 PGRN does not inhibit TNFR1 signalling. (a) BMDMs were pre-treated with human PGRN for 30 min and stimulated as indicated with TNF or LTa. Cell lysates were analysed by western blot for degradation of IkBa and phosphorylation of ERK, JNK and p38. (b) BMDMs were pre-treated with human or mouse PGRN or anti-TNFR1 and TNFR2 as indicated. Cells were then stimulated with TNF for 15 min and cell lysates were analysed by western blot for degradation of IkBa and phosphorylation of ERK. (c,d) U937 cells stably transfected with a NF-kB–GFP reporter were pre-treated with indicated PGRNs and stimulated with TNF or LTa. Twenty-four hours after stimulation, expression of GFP was assessed by flow cytometry. (e) U937 Cells were pre-treated with PGRNs followed by addition of TNF or LTa with or without SM compound for 48 h. (f) MDFs were pre-treated with human PGRN and then stimulated with TNF or LTa with or without SM for 24 h. Cell death was assessed by propidium iodide staining and flow cytometry. Error bars ¼ s.d.; n ¼ 3.
with a NF-kB–GFP reporter. Consistent with our previous results we did not observe any reduction in NF-kB activation after 24-h TNF or LTa stimulation by either of the three tested PGRNs (Figures 1c and d). Immunology and Cell Biology
PGRN does not inhibit TNF- or LTa-dependent apoptosis and necroptosis Loss of cellular signalling components such as inhibitors of apoptosis proteins (IAPs) converts the pro-survival signal of TNFR1 to a cell
Progranulin does not inhibit TNFR1 signalling N Etemadi et al 663
death-inducing stimulus.16 To test whether PGRN could interfere with TNF- or LTa-induced apoptosis, we pre-treated U937 cells with three different sources of PGRN at various PGRN-to-TNF ratios before stimulating them with either TNF or LTa in combination with smac mimetic (SM) for 48 h (Figure 1e). We did not observe any difference in cell death in the presence or absence of PGRN regardless of the TNF-toPGRN ratio (data not shown). Upon caspase inhibition, engagement of TNFR1 can result in alternative form of cell death, termed programmed necrosis (necroptosis).17 As an additional, independent read-out of TNF signalling in another cell type, we therefore examined the effect of PGRN on apoptosis and necroptosis induced by TNF or LTa in mouse dermal fibroblasts (MDFs, Figure 1f). In agreement with our previous results, apoptosis, where caspases are active, and necroptosis, where caspases were inhibited by pan caspase inhibitor (QVD), induced by TNF or LTa and SM were unaffected by PGRN treatment. DISCUSSION PGRN can function as an anti-inflammatory molecule; however, the mechanism for the anti-inflammatory action of PGRN is unknown. For this reason the report by Tang et al. demonstrating that PGRN directly inhibited TNF binding to, and inflammatory signalling from, TNFR1 and TNFR2 provoked intense interest. To extend the findings of Tang et al., we sought to determine whether PGRN could inhibit TNF/TNFR1 signalling differentially to LTa/TNFR1 signalling. However, in contrast to the results published by Tang et al. we were unable to observe any effect on either TNF- or LTa-induced signalling in various cell types in either acute or chronic signalling scenarios with three types of PGRN from two commercial sources. In our studies we used different cell types from both mouse and human to exclude cell type or species-specific effects. In contrast to the results from Tang et al., we were unable to observe any inhibition of TNFR1-induced signalling: the timing and degree of IkBa degradation and phosphorylation of ERK, JNK and p38 were all indistinguishable regardless of whether cells were pre-treated with PGRN or not. Whereas in the same experiments anti-TNFR1 was able to completely block these signalling events. Furthermore, TNF-induced apoptosis or necroptosis were not inhibited by any of the PGRNs tested. These results are consistent with those published by Chen et al.,13 who also tested a number of different PGRNs in a range of TNF-dependent assays but failed to detect any inhibition of TNF signalling. Our assays were performed under the assumption that the identity of our commercially sourced recombinant PGRN is correct, an issue that was discussed by Tang et al.15 We controlled this assumption using both western blot and mass spectrometry, and we confirmed that the recombinant PGRNs contained a single species that ran at the correct size that was detected with an anti-PGRN antibody and that contained tryptic PGRN peptides (data not shown). However, this analysis cannot determine whether the recombinant PGRN is correctly folded. However, Atsttrin, a recombinant molecule synthesised from three non-contiguous granulin domains, was reported to have the same biological effects as PGRN by Tang et al. This indicates that higher-order structure of PGRN is not critical for its purported activity. It was possible that high endogenous levels of PGRN in the cell culture supernatants could interfere with the action of recombinant PGRN. However, we did not detect any PGRN in cell lysates or supernatants of MDF and U937 cells (data not shown). BMDMs secreted low levels of PGRN in the supernatant; however, as the signalling kinetics were identical in all cell types tested and levels of endogenous PGRN significantly below the levels added ectopically,
it is unlikely that endogenous PGRN saturates TNF receptors preventing recombinant PGRN from inhibiting further. Our results support those described by Chen et al. and contradict those of Tang et al. Outside of the explanations described above we cannot explain why we were unable to repeat the observations of Tang et al. Regardless of the explanation, our results suggest that PGRN inhibition of TNF signalling is unlikely to be easy to observe with commonly available reagents and suggest that caution should be used before trying to develop reagents based on PGRN for clinical and preclinical studies. METHODS Cell culture and reagents All cells were maintained in Dulbecco’s Modified Eagle’s Medium þ 10% fetal calf serum. MDFs and BMDMs were generated using a previously described protocol.14 Stable NF-kB–GFP reporter U937 cells were generated using the NF-kB lentiviral reporter vector pTRH1 mCMV NF-kB dscGFP from System Biosciences (Mountain View, CA, USA). All tissue-harvesting procedures were performed according to the guidelines of the Animal Ethics Committee of the Walter and Eliza Hall Institute of Medical Research (WEHI).
Ligands and compounds Fc-TNF and Fc-LTa were produced and purified as described and used at a concentration of 10 ng ml 1 (for BMDMs, U932) or 100 ng ml 1 (for MDFs).18 The SM, Compound A, has been previously described,16 was synthesised by TetraLogic Pharmaceuticals and used at a concentration of 500 nM. Pan caspase inhibitor, QVD (10 mM; MP Biomedicals, Seven Hills, NSW, Australia) was added 1 h before TNF or LTa and SM treatment. Recombinant PGRNs were purchased from R&D Systems (Minneapolis, MN, USA) (human: 2420-PG-050) and Adipogen (San Diego, CA, USA) (human: AG-40A-0188; mouse: AG-40A-0080) and used at 250 ng ml 1. Monoclonal TNFR1 and 2 antagonistic antibodies were purchased from R&D Systems (MAB430; MAB4262) and used at a concentration of 2 mg ml 1.
Death assay and flow cytometry Cells cultured in 12- or 24-well tissue culture plates were harvested 24 or 48 h after stimulation with TNF or LTa/smac-mimetic/Q-VD-Oph and cell death was measured by propidium iodide staining and flow cytometry on a FacsCalibur (BD Biosciences, North Ryde, NSW, Australia). Data processing was performed using Weasel software (WEHI).
Western blotting and antibodies Cell lysates were prepared using DISC lysis buffer as described before,14 loaded on NuPAGE Bis-Tris gels (Life Technologies, Mulgrave, VIC, Australia) and transferred on to Immobilon-P PVDF (Millipore, Billerica, MA, USA) or Hybond-C Extra Nitrocellulose membranes (GE Healthcare, Rydalmere, NSW, Australia). Membranes were blocked and antibodies diluted in 5% skim milk powder or 5% bovine serum albumin in phosphate-buffered saline and 0.1% Tween20. Antibodies used for western blot: phospho-ERK1/2, phospho-JNK1/ 2, phospho-p38, p38, ERK1/2, JNK1/2, IkBa (Cell Signaling, Danvers, MA, USA), b-actin (Sigma Aldrich, St Louis, MO, USA). Signals were detected by chemiluminescence (Millipore) after incubation with secondary antibodies conjugated to horseradish peroxidise.
CONFLICT OF INTEREST JS is a member of the Scientific Advisory Board of TetraLogic Pharmaceuticals. All remaining authors declare no conflict of interest.
ACKNOWLEDGEMENTS This work was supported by the National Health and Medical Research Council of Australia (NHMRC) grants no. 541901, 541902, 602516, 1022916 and 1025594 and made possible through Victorian State Government Operational Infrastructure Support and Australian Government NHMRC Immunology and Cell Biology
Progranulin does not inhibit TNFR1 signalling N Etemadi et al 664 IRIISS (361646). UN was supported by the Swiss National Science Foundation (SNSF, fellowship #PA00P3_126249).
1 Zhou J, Gao G, Crabb JW, Serrero G. Purification of an autocrine growth factor homologous with mouse epithelin precursor from a highly tumorigenic cell line. J Biol Chem 1993; 268: 10863–10869. 2 Zhang H, Serrero G. Inhibition of tumorigenicity of the teratoma PC cell line by transfection with antisense cDNA for PC cell-derived growth factor (PCDGF, epithelin/ granulin precursor). Proc Natl Acad Sci USA 1998; 95: 14202–14207. 3 Toh H, Chitramuthu BP, Bennett HPJ, Bateman A. Structure, function, and mechanism of progranulin; the brain and beyond. J Mol Neurosci 2011; 45: 538–548. 4 Cenik B, Sephton CF, Cenik BK, Herz J, Yu G. Progranulin: a proteolytically processed protein at the crossroads of inflammation and neurodegeneration. J Biol Chem 2012; 287: 32298–32306. 5 Zhu J, Nathan C, Jin W, Sim D, Ashcroft GS, Wahl SM et al. Conversion of proepithelin to epithelins: roles of SLPI and elastase in host defense and wound repair. Cell 2002; 111: 867–878. 6 Kessenbrock K, Fro¨hlich L, Sixt M, La¨mmermann T, Pfister H, Bateman A et al. Proteinase 3 and neutrophil elastase enhance inflammation in mice by inactivating antiinflammatory progranulin. J Clin Invest 2008; 118: 2438–2447. 7 Bossu P, Salani F, Alberici A, Archetti S, Bellelli G, Galimberti D et al. Loss of function mutations in the progranulin gene are related to pro-inflammatory cytokine dysregulation in frontotemporal lobar degeneration patients. J Neuroinflammation 2011; 8: 65. 8 Martens LH, Zhang J, Barmada SJ, Zhou P, Kamiya S, Sun B et al. Progranulin deficiency promotes neuroinflammation and neuron loss following toxin-induced injury. J Clin Invest 2012; 122: 3955–3959.
Immunology and Cell Biology
9 Tang W, Lu Y, Tian QY, Zhang Y, Guo FJ, Liu GY et al. The growth factor progranulin binds to TNF receptors and is therapeutic against inflammatory arthritis in mice. Science 2011; 332: 478–484. 10 Zhao Y-P, Tian Q-Y, Liu C-j. Progranulin deficiency exaggerates, whereas progranulinderived Atsttrin attenuates, severity of dermatitis in mice. FEBS Lett 2013; 587: 1805–1810. 11 Kawase R, Ohama T, Matsuyama A, Matsuwaki T, Okada T, Yamashita T et al. Deletion of progranulin exacerbates atherosclerosis in ApoE knockout mice. Cardiovasc Res 2013; 100: 125–133. 12 Moelants EA, Mortier A, Van Damme J, Proost P. Regulation of TNF-alpha with a focus on rheumatoid arthritis. Immunol Cell Biol 2013; 91: 393–401. 13 Chen X, Chang J, Deng Q, Xu J, Nguyen TA, Martens LH et al. Progranulin does not bind tumor necrosis factor (TNF) receptors and is not a direct regulator of TNFdependent signaling or bioactivity in immune or neuronal cells. J Neurosci 2013; 33: 9202–9213. 14 Etemadi N, Holien JK, Chau D, Dewson G, Murphy JM, Alexander WS et al. Lymphotoxin a induces apoptosis, necroptosis and inflammatory signals with the same potency as TNF. FEBS J (e-pub ahead of print 2 July 2013 doi:10.1111/febs.12419). 15 Chuanju Liu J, Jian, Jyoti Mundra, Sardar Uddin, Elena Gonzalez Gugel, Brendon Richbourgh. Properly folded PGRN Can Bind TNFR. J Neurosci 2013, accessed 7 June 2013; Available from http://www.jneurosci.org/content/33/21/9202.full?sid=23c56273-8c72-4648-9fdd-2aad55f35b11- responses. 16 Vince JE, Wong WW, Khan N, Feltham R, Chau D, Ahmed AU et al. IAP antagonists target cIAP1 to induce TNFalpha-dependent apoptosis. Cell 2007; 131: 682–693. 17 Murphy JM, Czabotar PE, Hildebrand JM, Lucet IS, Zhang JG, Alvarez-Diaz S et al. The pseudokinase MLKL mediates necroptosis via a molecular switch mechanism. Immunity 2013; 39: 443–453. 18 Park B, Buti L, Lee S, Matsuwaki T, Spooner E, Brinkmann MM et al. Granulin is a soluble cofactor for Toll-like receptor 9 signaling. Immunity 2011; 34: 505–513.
